Coherent Diffraction Imaging for Enhanced Fault and Fracture Network Characterization Benjamin Schwarz1 and Charlotte M
Total Page:16
File Type:pdf, Size:1020Kb
https://doi.org/10.5194/se-2020-87 Preprint. Discussion started: 20 May 2020 c Author(s) 2020. CC BY 4.0 License. Coherent diffraction imaging for enhanced fault and fracture network characterization Benjamin Schwarz1 and Charlotte M. Krawczyk1,2 1GFZ German Research Centre for Geosciences, Albert-Einstein-Str. 42-46, 14473 Potsdam, Germany 2Technical University Berlin, Ernst-Reuter-Platz 1, 10589 Berlin, Germany Correspondence: Benjamin Schwarz ([email protected]) Abstract. Faults and fractures represent unique features of the solid Earth and are especially pervasive in the shallow crust. Aside from directly relating to crustal dynamics and the systematic assessment of associated risk, fault and fracture networks enable the efficient migration of fluids and, therefore, have a direct impact on concrete topics relevant to society, including climate-change mitigating measures like CO2 sequestration or geothermal exploration and production. Due to their small-scale 5 complexity, fault zones and fracture networks are typically poorly resolved and their presence can often only be inferred indi- rectly in seismic and ground-penetrating radar (GPR) subsurface reconstructions. We suggest a largely data-driven framework for the direct imaging of these features by making use of the faint and still often under-explored diffracted portion of the wavefield. Finding inspiration in the fields of optics and visual perception, we introduce two different conceptual pathways for coherent diffraction imaging and discuss respective advantages and disadvantages in different contexts of application. At the 10 heart of both of these strategies lies the assessment of data coherence, for which a range of quantitative measures is introduced. To illustrate the approaches versatility and effectiveness for high-resolution geophysical imaging, several seismic and GPR field data examples are presented, in which the diffracted wavefield sheds new light on crustal features like fluvial channels, erosional surfaces, and intricate fault and fracture networks on land and in the marine environment. 1 Introduction 15 Crustal faults and fracture systems are of significant importance for the structural interpretation of geophysical images. Result- ing from acting forces they not only encode past configurations of local stress fields, but also represent primary indicators of man-made or natural hazards or fluid flow in the subsurface (Sibson, 1994). In addition, the delineation of faults also helps to shed light on the mechanical properties of the host material and provides valuable assistance in tracking horizons and spatially linking stratigraphic units in sedimentary regimes. Crystalline-rock environments, which are of special interest for geothermal 20 exploration and production, are known to be brittle and scarred by intricate fracture networks, whose successful identification and characterization has an immediate impact on the desired transition to sustainable energies. Despite their importance, pro- nounced direct geophysical images of crustal faults, in particular when temporarily inactive, remain largely elusive, owing in large parts to their sub-wavelength structural complexity and the seemingly diffuse and complex wavefields that are typically associated with them. 1 https://doi.org/10.5194/se-2020-87 Preprint. Discussion started: 20 May 2020 c Author(s) 2020. CC BY 4.0 License. 25 With a long history in optical imaging, the wave process of diffraction is synonymous with the highest possible resolution achievable in a reconstruction (Born and Wolf, 2013). Large parts of the Earth’s crust are known to heavily diffract incoming seismic or electromagnetic radiation. However, exploration and earthquake seismology either rely on transmitted, reflected and converted arrivals or surface waves and often implicitly ignore weaker, seemingly uncorrelated contributions for the direct imaging of the subsurface. Constrained by the interference with other typically stronger reflected or transmitted phases, individ- 30 ual diffractions are often hard to identify on isolated records, despite the fact that they represent coherent signal. The suitability of seismic diffractions as a direct fault indicator was already explored in the 1950s and was further investigated in the follow- ing two decades (Krey, 1952; Kunz, 1960; Trorey, 1970; Berryhill, 1977). While these studies were mostly concerned with the accurate numerical modelling of the diffraction response, the first imaging attempts, despite their novelty, largely suffered from inadequate data quality (Landa et al., 1987; Kanasevich and Phadke, 1988). After an extended period in which seismic 35 migration and waveform inversion techniques evolved to their current sophistication (Etgen et al., 2009; Virieux and Operto, 2009), advancements in data acquisition led to a recent re-discovery of diffraction imaging for geophysical applications. Coherence is a collective property of a wavefield and can be viewed as a pre-requisite for migration-type imaging. Recent decades have proven the usefulness of systematically assessing this property for applications like noise suppression (Mayne, 1962), wavefront attribute extraction (Gelchinsky et al., 1999a, b; Jäger et al., 2001), data interpolation and regularization 40 (Baykulov and Gajewski, 2009; Hoecht et al., 2009), wavefield separation (Bergler et al., 2002), velocity inversion (Symes and Carazzone, 1991; Billette and Lambaré, 1998; Duveneck, 2004), or passive-source localization (Schwarz et al., 2016; Diekmann et al., 2019). With the increasing availability of dense acquisition systems, different forms of coherence arguments have been invoked in seismic and ground-penetrating radar (GPR) diffraction imaging. Arguably one of the most important applications and stumbling blocks for successful high-resolution imaging is the separation of the faint diffracted wavefield 45 from stronger, often heavily interfering contributions. While some approaches introduced a diffraction bias in the migration scheme (Khaidukov et al., 2004; Moser and Howard, 2008; Klokov and Fomel, 2012), other strategies aim at extracting the weak diffraction response in a separate step before imaging (e.g. Bansal and Imhof, 2005; Fomel et al., 2007). Likewise applied before migration, there exist techniques that make direct use of wavefield coherence for diffraction sepa- ration (Berkovitch et al., 2009; Dell and Gajewski, 2011; Bauer et al., 2016; Bakhtiari Rad et al., 2018). While these methods 50 specifically target the diffracted wavefield for extraction, recent developments have shown that a more surgical, amplitude- preserving separation can be achieved by assessing the coherence of reflections instead (Schwarz and Gajewski, 2017a; Schwarz, 2019b). Although other methods like e.g. plane-wave destruction can achieve a similar quality of extraction in many applications, the systematic and physically intuitive assessment of coherence can be carried out in any imaginable data config- uration and allows for a seamless integration of data enhancement, wavefield separation, and imaging into a single framework. 55 Recent studies suggest that, owing to their unique properties, diffractions also lend themselves well for velocity inversion (Sava et al., 2005; Fomel et al., 2007; Decker et al., 2017; Bauer et al., 2017). These approaches bear the potential for a self- contained imaging and inversion cycle that is also applicable in the case of offset-limited acquisitions as they can often be found in academia (Preine et al., 2020). 2 https://doi.org/10.5194/se-2020-87 Preprint. Discussion started: 20 May 2020 c Author(s) 2020. CC BY 4.0 License. With only few exceptions (e.g. Landa et al., 1987; Heincke et al., 2006; Dell et al., 2019), the potential of quantitative 60 coherence measures for directly forming noise-robust, contrast-rich images remains largely unexplored. Building on recent advances in adaptive processing and weak-wavefield enhancement, we present a strategy for seismic and GPR diffraction imaging that makes direct use of wavefield coherence for scatterer detection. After a brief elaboration on typical characteristics and unique properties of diffraction phenomena, we introduce two different means of reconstructing a scatterer with the help of coherence measurements. Underpinning both these pathways we introduce generalized coherence measures and systematically 65 investigate their tolerance with respect to imperfect, i.e. noisy, sparse, or incomplete data and make suggestions with respect to their applicability. Concluding community-spanning seismic and electromagnetic examples suggest that coherent diffraction imaging not only leads to overall highly-resolved subsurface reconstructions, but also directly and reliably highlights small- scale erosional features, faults and fractures. 2 Wave diffraction 70 Diffraction can loosely be defined as a waves ability to enter shadow zones, which are forbidden regions in geometrical optics. More precisely, diffraction occurs when a wavefield encounters a relevant property change that has a local curvature of or below the wave length (Born and Wolf, 2013). Thus, diffraction is a scale-spanning phenomenon that is only predicted and fully captured in a wave theoretical framework. To provide some intuition, Figure 1 illustrates some of the key properties of diffractions and how they can be of use for geophysical subsurface imaging (for more details, see Schwarz, 2019a). As arguably 75 the first rigorous experimental evidence, Young’s slit experiments concluded that when light hits